PROJECT SUMMARY The diagnosis of Bipolar Disorder (BPD), a heritable disorder characterized by periods of depression and mania, is extremely challenging in clinical practice and takes on average 5-10 years. One of the main reasons for this delay is that the majority of patients first manifest depressive symptoms (depression-first BPD), often resulting in misdiagnosis with major depressive disorder (MDD). Misdiagnosis can lead to poor clinical outcomes, increased burden for patients and high healthcare costs. The goal of this proposal is to identify genetic and phenotypic signatures that distinguish depression-first BPD from MDD, in order to accelerate BPD diagnosis. During the mentored phase (K99) of the award, the PI will perform a genome wide association study to characterize the genetic architecture of differences between depression-first BPD and MDD, using available data from the collaborative Psychiatrics Genetics Consortium (Aim 1a). In a Dutch BPD cohort (n=1,750 cases) she will use machine learning methods to identify molecular, neuro-cognitive and behavioral profiles to classify depression-first BPD versus mania-first BPD (BPD with first onset of mania) and identify causal relations among predictive features (Aim 1b). During the independent phase (R00) of the award, the PI will study BPD diagnosis in a genotyped and extensively phenotyped Colombian cohort of ~5,000 cases with severe mood disorders, for which high quality electronic health records (EHR) have been collected since 2005. She will i) predict if and when patients will switch from MDD to BPD diagnosis, ii) distinguish BPD from matched MDD patients, and iii) classify BPD patients into depression-first and mania-first categories (Aims 2 and 3). She will estimate heritability of identified signatures and integrate results from both Dutch and Colombian cohorts. Building upon her expertise in complex trait genetics and method development, this award will allow the PI to extend her knowledge-base of clinical phenotyping in psychiatry; grow her expertise in statistical modeling; and strengthen her communication, mentoring, and leadership skills to prepare for a successful career as an independent researcher. The PI will complete the K99 phase of the award at UCLA, which fosters an ideal training environment. The strong advisory committee with Drs. Freimer, Ophoff, and Eskin (UCLA) and Dr. Sabatti (Stanford) will facilitate her scientific and personal development and promote her long-term career goal of tackling clinically relevant problems in psychiatric illness through the study of genetic, neuro-cognitive and behavioral mechanisms. The PI will first focus on the diagnosis of BPD, and then expand, in the future, into other psychiatric trait-mechanisms, such as disease severity and drug response. This award will be fundamental in support of her career goals. Successful completion of the specific aims of this proposal is likely to significantly improve the time to BPD diagnosis, and thereby reduce the burden of this debilitating disorder for patients and society. !